Fewest gain stages.....

[dave251]

What are the fewest gain stages that can be used to build a "practical" (ie, something strong enough to be heard in a busy nightclub) amp? Will solid state require more gain stages? Or is the pure, single ended class A tube amp the simplest?

[mooreamps]

What are the fewest gain stages that can be used to build a "practical" (ie, something strong enough to be heard in a busy nightclub) amp? Will solid state require more gain stages? Or is the pure, single ended class A tube amp the simplest?

I would say one pre-amp tube with two gain stages inside and one power tube, either an EL34 or a 6L6 for about 12 watts single ended.

-g

[Steve Conner]

It depends on whether you want to be heard with a big clean tone, or a dirty screech. Also how loud the rest of your band are, and what genre of music you're playing.

The classic amps had two preamp stages (the two halves of a 12AX7) with the tone stack in between attenuating it somewhat again. Then there was a phase inverter and two power tubes. That's four gain stages. Mooreamps is essentially recommending three stages. Single-ended amps like the Fender Champ and Epiphone Valve Jr. get by fine with three, but I would like another stage to really be able to beat on bigger power tubes like the 6L6.

I built an experimental 60 watt amp with a switchable choice of 3, 4 or 5 gain stages before the phase inverter. The first stage also has an adjustable Rk to control the gain from a little more than 1, up to the usual 50 or so. Therefore the amp can have the equivalent of anywhere between 2 and 5 stages before the PI. It can be heard in a busy nightclub regardless of the setting. I know because I've played several gigs with it now.

Of course 2 stages requires a higher master volume setting and gives a clean sound, but it's still usable for tunes that require clean guitar. It also makes a good bass amp on this setting. 5 stages give a deafening doom-laden grind with squealing feedback, that I rarely get to use in a live setting.

[dave251]

WE have done three prototypes, the first two being single ended....but, even those two had four gain stages. The first one used a WE300B, the second would work with a 6L6, EL34, or 6550. I preffered the EL34. And, while we could drive a super efficient JBL two way cab(E130 and baby cheeks tweeter) up to club gain levels(clean), the cabinet was HUGE, and VERY expensive.

The third prototype is based on an AB design using a pair of EL 34s on the output. Preamp is a 12AX7, with a 12AT7 phase splitter...single knob tone stack(treble rolloff). That's four gain stages. We're getting a very clean 30 to 35 watts from the amp. BTW...this is an "acoustic/electric" amp, using an Em Beta 12LT for a speaker.

But my question is, can a solid state amp be done with as few gain stages? From a passive guitar input through output? I have tried just about EVERY solid state "acoustic guitar" amp, and ours is definitely "sweeter" sounding with my guitars....not as "accurate" maybe......

I'm wondering if I can get that tube "clarity" with a low feedback SS circuit, and the fewest gain stages. I believe that too many gain stages squash the "feel" of the guitar.

[Bruce / Mission Amps]

One tiny FET, a 12AX7 preamp tube, two 6V6 power tubes.

Input jack into a flat response, small signal FET preamp, gate held up with 1M resistor and the drain running with a well filtered +16vdc rail.
100mvac input.... +2vac to 3vac output.
The output of this FET RC coupled into a standard Fender brownface Deluxe volume tone control circuit.
1M audio volume pot, bypassed with a, 1M audio pot tone control, 500pF treble cap and grounded .01uF "bass" cap.
This audio recovered out of the volume pot wiper with a 12AX7 triode, running +30dB gain, RC coupled to another 12AX7 triode which will be wired as a cathodyne driver, pushing two push pull fixed biased 6V6s at +400vdc and an 8K output transformer into a single 12", 100dB speaker.
Power supply a 300vac@120ma PT with solid state rectifier.
Should make well over 15-18 watts and be very lovely.
Bigger amp?
Two 6L6s with bigger transformers and you shoudl get well over 30 watts.

[Amp Kat]

That sounds really nice Bruce and have often wondered about Fets since they do have tube like curve characteristics. Even though I've messed with them before mainly in mic pres and with a servo to match sides of duals they seem to be a good fit for guitar application. Which one did you use and does it sustain pretty long ? Certainly Dave is looking for something simple which this is with the right tranny for the low voltage fet supply. The SE is pretty basic also but may not be quite loud enough unless you designed for maybe a 6L6, 6550 or KT-88. I like the 2-6v6 idea with a tube rectifier. Maybe a Vibrochamp with 2 -6v6's would be nice.

[mooreamps]

One big power tube

Well, one could push a 6L6 or an EL34 to 18 watts SE. Maybe be a little hard on the power tube. For anything more than that, and still wanting the fewest amount of gain stages, one could go with an RCA 813 Beam Power tube running SE for 45 watts.

-g

[Bruce / Mission Amps]

I've tried a bunch... parallel 6V6s in SE class A, a SE 6L6GC, a SE 6550, SE KT88.
The last 6L6GC SE class A amp I built used a Hammond 125FSE, 20 watt output tranny but was for a small harp amp.
I was expecting a lot more output but it still didn't make much more then about 13-15 watts either.
The 6550/KT88 with a 130ma PT and 200ma choke input filter sounded great but still coax much more then about 16-17 watts output of cleanish power using the same iron.

I think it just must take a ton of current, lower plate voltages and a huge output transformer, wound exclusively for SE class A amp designs.

They are fun to mess with but the best sounding one I've built has probably been done to death all over the world.
Basicaly a glorified Fender tweed Princeton.
Mine used a single NOS 6V6GT, an old Thordarson 300v@120ma PT, a NOS 5Y3GT rectifier and a simple 10 watt Schumacher SE OT.
It made about 5-6 watts clean and I used to call it the Soulkicker Jr.

[Bruce / Mission Amps]

That sounds really nice Bruce and have often wondered about Fets since they do have tube like curve characteristics. Even though I've messed with them before mainly in mic pres and with a servo to match sides of duals they seem to be a good fit for guitar application. Which one did you use and does it sustain pretty long ? ...

I use a J201 at 16vdc to 20vdc.
The amp is a push pull 6G2 Princeton type amp but with a FET first preamp and it is super simple.
The +16vdc for the FET is derived from a voltage divider after the last filter cap, I think something like 180K and 10K, with a little 22uF@50v filter cap across the grounded 10K resistor.
I've also used a 1/2w 18v zener to ground after the large voltage dropping resistor... the parallel FET draws so little current that it very easy to get the voltage where you want it.

[Steve Conner]

Dave251, gain stages don't really have the same meaning in solid-state design, because the gain of a "stage" can vary very widely depending on how you make it.

Indeed, you can quibble endlessly over how to define exactly what is a "stage." For instance, you could conceivably make a fairly powerful guitar amp using only a LM12 80 watt power op-amp:

http://www.national.com/pf/LM/LM12CL.html

and a handful of passives. It has an open-loop gain of around 30,000. So, all the gain you need... in one stage! (Or is it really one stage? How is the LM12 constructed internally?)

A LM12 power stage modified for soft clipping, hooked to a single JFET preamp, with a Fender tone stack and volume in between, might actually sound quite good for acoustic guitar. Since it's an op-amp, I'm sure you can also cook up presence and resonance controls that increase its output impedance in efforts to make it sound more "tubey".

PS: I ROFL'd at the "baby cheeks tweeter" I think I know the JBL tweeter you mean.

[MWJB]

Dave251, I'm not sure that just reducing the no of gain stages is the answer (especially regarding your El34 AB design you mention). As others have stated, there's no easy way around a typical 4 stage set up in P-P (if you want to stick to all tube that is). Looking at how/why the stages you have distort/squash might be better (do you have a schem? Is a single tone pot and 12AX input the way to go?).

[jrfrond]

Here's something that works beautifully:

EF86 pentode preamp-volume control w/bright cap-12AX7 LTP PI w/no NFB-Post PI dual pot MV-Vox-style "BRILLIANCE control (as the sole tone control)-pair of cathode-biased 6550's.

You'll get anywhere from clean to cranked with this. I am listening to my son rehearse on it as we speak. This is the format I chose when I built custom amps 10 years ago. I call it "Tweed on Steroids". Very responsive and lots of tone due to the short signal path and low gain loss with no tone stack. If you need extra EQ, you can use a box ahead of it.

[MWJB]

EF86 also gives me an idea re. the original post/fewest stages - EF86 input, straight into a self-split 4xEL84 output...less efficient than a PI driven push pull output, maybe 20W output (no guarantees, just putting it out there as a concept - as opposed to John's tried & tested amp)? Swings & roundabouts though, expensive transformers vs cheaper transformers & an extra tube & a PI driven, fixed bias, pair of EL84 would probably do the same job. Then again this would just be an exercise in fewest stages, in terms of parts count a 2 stage preamp, based on a twin triode, into the self split output would use the same number of tubes.

[jpfamps]

Here's something that works beautifully:

EF86 pentode preamp-volume control w/bright cap-12AX7 LTP PI w/no NFB-Post PI dual pot MV-Vox-style "BRILLIANCE control (as the sole tone control)-pair of cathode-biased 6550's.

You'll get anywhere from clean to cranked with this. I am listening to my son rehearse on it as we speak. This is the format I chose when I built custom amps 10 years ago. I call it "Tweed on Steroids". Very responsive and lots of tone due to the short signal path and low gain loss with no tone stack. If you need extra EQ, you can use a box ahead of it.

I think of the DrZ amps use this gain structure, ie EF86, phase splitter, power valves.

[jrfrond]

I think of the DrZ amps use this gain structure, ie EF86, phase splitter, power valves.

Hey, I didn't invent it, I just plagarized it! Just have a look at the schematics for AC4, AC10, AC15 or even Matchless' DC30 pentode channel (minus the bass rolloff rotary switch). Again, none of this stuff is rocket-science. I just decided to scrap the oh-so-typical EL84 power section for 6550's cathode-biased, with a couple of minor tweaks. As it turns out, the tightness of the EF86 works well for higher-power amps, because it seems to control "bloom" better than triode designs.